AI in Hypersonic Flight Control (Adaptive RL for Stability & Safety)

Aim

Course Objectives

Syllabus

📅 Day 1 – Safety, Ethics & High-Speed Flight Basics (Non-Weaponized)

• Where AI fits in safety-critical aerospace: hazard analysis, “human-on-the-loop,” and oversight.
• High-level aerothermodynamics concepts (no design details): why high speeds amplify uncertainty and sensing challenges.
• Assurance artifacts: requirements traceability, safety cases, and model/system cards.
• Hands-on: Build a policy-aware requirements matrix and hazard log for a benign high-speed test article (e.g., generic aero model) focusing on oversight, abort criteria, and geofencing.

📅 Day 2 – Robust Perception & State Awareness (Conceptual)

• Sensor integrity at high dynamic pressure: fault concepts, latency awareness, and graceful degradation (no algorithms).
• Observability at a glance: what it means to estimate states safely without revealing implementation.
• Validation & test planning: scenario coverage, limits, and transparency for operators and regulators.
• Hands-on: Draft a verification & validation (V&V) plan: define test envelopes, safety monitors, and operator intervention thresholds for a conceptual high-speed vehicle.

📅 Day 3 – Governance, Certification & Human Oversight

• Standards and certification thinking for AI components (conceptual): documentation, audits, incident reporting.
• Envelope thinking without controllers: defining stay-out zones, rate limiters, and conservative defaults.
• Human-machine interfaces: alerting, explainability for operators, and abort workflows.
• Hands-on: Assemble an assurance case outline (safety case) with roles, evidence to collect, and an operator SOP for safe testing and shutdowns.

Who Should Enrol?

Course Outcomes